Talk Next Generation of Carcass and Accident Reporting Systems
  • Fraser M Shilling
    Bio statement : I am the co-director of the Road Ecology Center at the University of California, Davis. I driect research and education programs at the REC and within the Department of Environmental Science and Policy at UC Davis.

    Reviewing interests: roadkill, road effect zone, wildlife crossing structrues, camera traps, monitoring, connectivity, traffic noise, sea level rise, sustainability.
    Country : US
    Contact : fmshilling@ucdavis.edu
    Website : http://roadecology.ucdavis.edu

  • David Paul Waetjen
    Bio statement : I am the lead web-programmer for the Road Ecology Center. i also develop websites to track species invasion, air and water quality, and sea level rise.
    Country : US
    Contact : dwaetjen@ucdavis.edu
    Website :
  • Kathryn Harrold
    Bio statement : I am a contract-researcher with the Road Ecology Center specializing in monitoring wildlife movement and fatalities in relation to highways and roads.
    Country : US
    Contact : kharrold111@gmail.com
    Website :
Abstract

Large-extent Carcass and Accident Reporting Systems (CARS) have a fairly typical set of goals and methods, which help in standardizing methods of data collection, analysis, visualization, and use in decision-making across systems. Most systems around the world collect carcass and/or crash data to increase the understanding of degree and geography of ecological impact, risks to drivers, placement and effectiveness of mitigation structures. Most systems also follow a common rubric for data collection of: where, when, what (species), who (observer), how, why and "other", where the latter category refers to characteristics of the roadway, habitat, or carcass. Most systems also use web-systems and databases to collect, store, and share mapped and tabular data. We report here on two major types of advancement in the collection and uses of CARS data.

The California Roadkill Observation System (CROS, http://wildlifecrossing.net/california) uses a form-based data entry system to record observations of carcasses that result from wildlife-vehicle collisions (WVC). Operating since 2009, it currently (Jun, 2016) contains >1,200 users and ~50,000 observations of >400 species of ground-dwelling vertebrates and birds, making it one of the most successful examples of crowd-sourced, roadkill data collection. The system can be used directly on a smart-phone screen, via a browser, making an app unnecessary. We have used data from this system to determine locations of high-density and/or statistically-significant clustering of WVC. These hotspots are potential locations for mitigation actions to reduce risks to drivers and wildlife. We have also used the data to inform two ecological studies: a state-wide study of mammal invasion and assessment of the ability of landscape "linkages" models to predict wildlife movement. Finally, volunteer-observers are carrying out studies to help us estimate total roadkill impact on species: collecting observations on transect roads and re-observing roadkill to estimate disappearance rates of carcasses from roads.

The newer California Highway Incident Processing System (CHIPS) scrapes publicly-available crash reports from the California Highway Patrol (CHP) incident website and stores them in a local database. The user can query and tag incidents for involvement of animals, type of accident (e.g., collision vs. avoidance), and outcome of accidents for driver and animal. After 11 months (Feb, 2015 - Jan, 2016), there were 1.23 million total crash records in the system, with ~8 incidents/day involving mule deer. About 1/3 of these resulted in the deer getting injured, but not dying immediately. A similar proportion involved a vehicle swerving to avoid an animal in the road and subsequently crashing. These accidents were more likely to involve significant vehicle damage and injury than collisions with animals. Data from this system will be used to estimate significance of WVC avoidance as a type of human-wildlife conflict that threatens driver safety, to estimate injury and property damage rates from all types of crashes involving different animal species and vehicles, and to compare crash hotspots with carcass hotspots. Because the system operates in real-time, we are also investigating the possibility of a continuously updated smart-phone application or Twitter feed to inform drivers and others of animal-related safety issues on highways.

The combination of crowd-sourced carcass data, agency-collected carcass data, and agency-reported crash data allows for a comprehensive and real-time view of human-wildlife conflict on roads. This means that the system provides decision-support from the individual driver scale to the state scale and from an historical perspective to what is happening right now.

Keywords
traffic mortality, accidents, wildlife-vehicle collisions, human-wildlife conflict, wildlife-vehicle conflict